FOICP-Miner: An Interactive Spatial Pattern Recommendation System Based on Fuzzy-Ontology

Spatial pattern mining is essential for the exploratory analysis of spatial data, with numerous efficient systems available that can discover various types of spatial patterns in large datasets. However, identifying patterns that are genuinely interesting to a specific user remains a significant challenge. To address this issue, we develop FOICP-Miner, an interactive system based on the fuzzy ontology. It is designed to facilitate the effective discovery of personalized spatial patterns tailored to individual user interests. FOICP-Miner employs the domain-specific fuzzy ontology to encapsulate users’ background knowledge. Users are then prompted to express their preferences on sample patterns. Finally, the system leverages a fuzzy ontology model to evaluate the users’ prior knowledge and extract their preferred patterns from the candidate pattern set. Our evaluation results demonstrate that FOICP-Miner identifies user-preferred patterns more effectively compared to the state-of-the-art researches.